Fig. 2

Optimal settings algorithm description. The following steps describe the algorithm in the block diagram: 1. The interface input includes the choices of sample type (sparse/dense) and optimization parameter (CNR/Imaging time); the imaging parameters like imaging time/PDT or SNR, flyback portion of the time, pixel size, number of pixels, and object size; filter order range for optimization; powers applied on the sample; calibration constants and beam diameters; 2. A range of TCs is set as an input for the next step for each filter order. The TC is relative to an arbitrary (pre-selected) PDT. The input parameters are passed to the next step; 3. The object is constructed (top hat or square wave normalized to 1). The SRS focal spot (from beam radii) is convolved with the object. The convolved object serves as an input for the RC filter response, for the TC parameter range. This range is with a course precision; 4. The filter response over the object is given to represent the signal propagation: For a single particle, the maximum of the response is determined as the signal (in arbitrary units). For a dense sample, the difference of the second local maximum and second local minimum is determined as the signal (in arbitrary units); 5. The noise is calculated with a TC and filter order range, relative to an arbitrary PDT. Together with the signal (from the normalized response), a CNR range is determined and the maximum of the CNR is found. The maximum CNR is then used to determine the optimal TC (relative to the arbitrary PDT). A smaller range with higher precision (smaller intervals) around this optimal TC from the last step is used to repeat steps 2–4. This gives the optimal TC (relative to an arbitrary PDT); 6. From the smaller TC range the noise is calculated relative to an arbitrary PDT, a CNR range is determined and the maximum of the CNR is found. The maximum CNR is then used to determine the optimal TC with a higher precision (relative to the arbitrary PDT) for each filter order; 7. If CNR is optimized, then the PDT is given and the TC is scaled with the given PDT and object size. The CNR is scaled with the calibration parameters, beam powers and TC. If imaging time/PDT is optimized, then the CNR is a given and the TC is extracted from the ratio of the given CNR and the maximum CNR from the previous step. The PDT is extracted from the arbitrary PDT and the TC ratio; 8. The response curve is plotted; 9. The imaging time is calculated (can be done directly from the input in step 1, or after the program runs, depending on the optimization choice)